How much is too much for our doctoral students and postdocs?
Have these pressures been the same for every generation of postdocs and PhD students in the past.....an advisor's perspective
(Updated 25.7.2023)
Reports on the prevalence of mental health problems in academia keep coming in, and it appears this is currently a widespread issue among doctoral students and postdocs. One reason for this is probably the heightened awareness of an importance attached to people’s mental well-being, compared to how it was in the past - obviously a very positive and necessary shift.
But is there more to it than this than greater sensitivity to the problem moving the baseline of ‘detection’ of mental health problems; in other words, is there something that has fundamentally changed in terms of expectations, workload, information processing, or other factors, compared to how it was for previous generations of early career researchers?
I think this is a very difficult question to answer conclusively, perhaps it’s even impossible; but I offer some thoughts here, and would love to hear from you about your take on this, whether you are a doctoral student, postdoc or a more senior researcher. Most of what I write about here, from the perspective of an advisor, applies to the field of ecology, but likely also beyond.
Information load and density (publications). I believe one of the most obvious parameters that has changed compared to the past (like, when I started my PhD) is the density of information in terms of published reports. While previously, it was often pretty clear who in the world was working on what, now the playing field is much larger, and many more groups will work on any one problem or issue. As a result, it is hard to keep up with the resulting torrent of information. There was of course always a lot to read for PhD students. But I think the pace of newly appearing information is unprecedented, and this literature is on top of the already existing body of classical papers.
Increased expectation for data visualization and concept figures. I have observed a steep increase in the general expectations for the presentation of figures for manuscripts, both for data and concept figures. For example, data figures now contain a lot more elements of interpretation. This is of course good, because it enhances the accessibility and attractiveness of figures. There certainly are also better tools for this than what used be available, like packages in R, or graphics software; but this is still yet another skill to acquire as a graduate student.
Statistics proficiency. Many statistics papers now read like stats exercises, at least the Methods parts. The level of statistical sophistication has also been increasing at a dizzying pace. Of course we need this, since better methods allow us to separate the signal from the noise, and especially in observational studies we need to use the best statistical tools available. Statistics is also quite a hurdle to some who are not primarily interested in data science, and it is definitely a whole bundle of techniques to learn to apply and to understand. R is the tool of choice in ecology. Machine learning is becoming more important in addition to classical tools. In some cases, like in molecular ecology, there are also ever-increasing improvements to bioinformatics pipelines that one needs to master in addition to statistics. Of course ecologists always had to deal with data and statistics, this is hardly new. But it seems the level of sophistication is way beyond what was available when I was in grad school. Again, a lot to ask of early career researchers who need to constantly keep up-to-date on these techniques.
Massive real-time interactivity through social media. Of course it is important for early career researchers to be present on the web, including on social media, as this is also an effortless way to make connections with people, to learn about ongoing work, and to receive inspiration. On the other hand, being bombarded with everybody’s work and success stories can induce FOMO, anxiety and feelings of inadequacy. Previous generations of researchers did not have the opportunity to engage in this social media ecosystem, but they were also not exposed to these potential downsides and this rapid-fire input.
Outreach/ communication. Engaging in communication of research results is potentially very rewarding and can offer exciting opportunities. However, these activities, as important as they are, can also be a huge time drain. While I applaud all postdocs and PhD students engaged in outreach, and good science communication is definitely becoming increasingly important in our current times, this work can add additional burdens on people.
Entirely new classes of tools (generative AI). These are times of fast-moving developments that could not have been foreseen by the average scientist even a year ago. Generative AI, something that most people never heard of before end of 2022, is now offering incredible opportunities, but also completely new challenges and also risks. Perhaps this is quite comparable to some other introductions of technology in the past, like availability of personal computers and the internet, which have also revolutionized the way we do science. But this just adds to all the other pressures.
So, what can we do?
Well, early career researches simply do need to learn and apply state-of-the-art techniques, there is no way around this. And they need to connect with other researchers. But how can this be done best?
People need the proper support to get this done within the limited time frame of a PhD or a postdoc. Mentorship and co-learning in supportive teams are going to be central for this.
Maybe from the beginning we need to prepare incoming early career researchers for the demands that they will face, such as a torrent of information in terms of publications, and steep learning curves for software, methods and stats.
When we devise new graduate programs we tend to load them up with interesting experiences (like summer schools, regular meetings, workshops, internships), and while this is good, it may also serve to add to the time pressures. Perhaps we need to have simpler, leaner programs focused on the basics? Everything else can be offers rather than program requirements.
Social media training is typically not part of the curriculum. Perhaps some basics should be included, or at least this should be discussed in lab meetings or workshops. There is much one can do to make that experience a better one.
Did I miss any important points, both in terms of novel pressures and the potential solutions?
What do you think of all this? Please let me know in the comments.
Great article, Prof. Rillig!
I agree with every detail in the article. A successful career is not necessarily correlated with the quantity/quality of publications. Research grants are highly competitive. Many of us, including myself, suffer from overdue unpaid fellowships (around three years). As time goes on, the situation only becomes more demanding.
Although I largely agree with everything you wrote, about the changes taking place, I do think it's important to point out that this perspective was written almost entirely from a PI perspective/agenda. This is not so strange, given your position of course. However, I cannot emphasize enough that a PhD, as well as a postdoc is also about furthering the respective candidates' own careers, not just about strengthening the PI's academic position. They have a completely different agenda and perspective, and the potential frictions in expectations originate from this misalignment, I think. In many places there's place to work on this (in theory) - as long as you do this off work time (in practice). And this is of course not explicitly stated anywhere by most PIs, but the extremely high expectations on work load make it so.
I speak from a 3+3 type postdoc position, where - even though there's room to develop personal things should I want this - my standard work load is 40+h-high already. Fitting in additional personal development stuff, courses, trainings, etc. would mean I'd have to do substantial overtime to make it through my day-to-day tasks.
I think it is also a PI's responsibility to keep in mind both perspectives, and also be ruthless in promoting healthy work-life balance, ie 40h work weeks where there's room for personal and PI agendas without too much overtime. (This very much a note to self as I transition into PI role myself)